IEEE 802.11/D2 Letter Ballot Response€¦ · Web viewResolutions for Comments on Clauses 12-16. 12.3.2 PMK e “...is seperated into to sublayers:...” “is separated into two

March 1997 doc.: IEEE P802.11-96/157R1Seq.

#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote

Comment/Rationale Recommended change Disposition/Rebuttal

Results of LMSC Ballot on Draft Standard 802.11 D5.0 -

Resolutions for Comments on Clauses 12-1612.3.2 PMK e “...is seperated into to sublayers:...” “is separated into two sublayers:” accepted as result of accepting

REVSEC9.DOC12.3.3 KC T It would be better layering if the standard included a

section here such as "(3) Service primitives that support timekeeping." and all timers moved out of

the MAC layer and into the PHY clause 12 as services. This would allow an implementation of the

entire MAC layer as an object that could be completely tested at this boundary with simulated

events.

postponed for full working group plenary

full plenary motion to reject the comment (Wim/Anil)

this change will be somewhat widesweeping and will certainly

cause delays in producing the draft.

(this comment is rejected by 20-0-7 vote)

12.3.3 TLP e Change first sentence to be literate English. Change to read “The primitives associated with communication between

the 802.11 MAC Sublayer and the 802.11 Physical Layer fall into two basic

categories:”

accepted as result of accepting REVSEC9.DOC

12.3.4.3 TLP e Put the two primitives PHYDATA.request and PHYDATA.indicate on separate lines within a single table

entry (as shown in the submitted revision-marked files).

Change column title to “Associated Primitive” and make a two-line entry in

the first data row, second column.


12.3.5.1.2 TLP E Yes Other portions of this standard use the syntax 0xNN for the hexadecimal number NN. This section uses 00

through FFh. Either syntax is acceptable, though the 0xNN syntax is more self-explanatory. But whichever, is used, please be consistent throughout the entire standard.

Use hexadecimal nomenclature consistent with the rest of this draft

standard.

postponed for full working group plenary

a vote in the full plenary approved using xxh nomenclature

throughout the standard as this method more formally matches

the use of SI units elsewhere in the draft.

WG vote (14-3-13)12.3.5.10

.2TLP e poor conceptualization and wording. For example, a

“channel assessment” process should observe a “channel”, not a “medium”.

Change second indented paragraph to read “The STATE parameter can be one

of two values: BUSY or IDLE. The


Results of LMSC Ballot D5.0 - Comments resolution 12-16 page 1 Vic Hayes, Chair, Lucent Technologies


#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


parameter value shall be BUSY if the channel assessment by the PHY sublayer

determines that the channel is not available. Otherwise the value of the

parameter shall be IDLE.”12.3.5.10

.3TLP e Use of undefined term. Change “clear” to “idle” twice. accepted as result of accepting

REVSEC9.DOC12.3.5.1

0.214.3.3.2

.29.2.5.2,

SB t N Clause 14.3.3.2.2 says:

The appropriate CS/CCA indication shall be generated prior to the end of each 50 µs slot time with the

performance specified in subclause Error: Referencesource not found (PMD).

(The CS/CCA indication is by PHYCCA.indicate as in figure 68)

While clause 12.3.5.10.2 says about PHYCCA.indicate:

This primitive shall be generated every time the status of the channel changes from channel clear to channel

busy or from channel busy to channel clear.

Clearly there is some conflict here - one says that the primitive is issued on a time basis once per slot time

even if the channel state has not changed, the other on a physical event (a change of channel state) irrespective of time. If I look at the PHY chapters the FH chapter

(Figure 68) would seem to follow 14.3.3.2.2 and the DS (Figure 83) follows 12.3.5.10.2

Actually this is pretty important for compliance given the rules that define when the back-off timer may, or

may not be decremented in 9.2.5.2

Correct conflict one way or the other - do I get a regular PHY CCA

indication per slot time, or only when the channel state changes.

(It also occurs to me that the first two sentences of clause 14.3.3.2.2 are

duplicated in the immediately previous clause.)

Clause 14.3.3.2.2 will be changed “shall be available ...”

clause 12 will be left unchanged

the MAC (9.2.5.2 and 9.2.4) backoff in whole slot increments,

and as long as the phy reports changes in CCA state

accordingly, not a problem. Taken as whole, the PHYs will report PHYCCA.indicate at the

specified times and can appear as continuous to the MAC

FG vote (9-0-3)

12.3.5.12.2

TLP e State machines do not “think”. Please avoid anthropomorphizing equipment and software.

Change first two sentences of second indented paragraph to read “The

RXERROR parameter can convey one or more of the following values:

NoError, FormatViolation, CarrierLost,




#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


or UnsupportedRate. A number of error conditions may occur after the PLCP's

receive state machine has detected what appeared to be a valid preamble and start

frame delimiter.”12.3.5.12

.2TLP e Use of inappropriate word. Change last word from “encountered” to

“detected”accepted as result of accepting

REVSEC9.DOC12.3.5.8.4 TLP e Yes The word “packet” (a network layer concept) is used

where “frame” is appropriate. Please use the appropriate OSI Basic Reference Model terminology.

Change “packet” to “frame”. accepted as result of accepting REVSEC9.DOC

12.all TLP e Use of undefined jargon Replace “node” with “station” (or “STA”) everywhere


12.all5.1.1.2 (c)

5.2.4.15.4

9.2.114.all

15.some16.all

TLP e Yes The wireless medium is definitely singular (unless there is an alternate universe with multiple “ethers”), or unless P802.11 is extending its charter to acoustic modes of

transmission.

change “edia” to “edium” everywhere except when referring to wired media.

for clauses 12 and 15


andREVSEC12.DOC

13.1.1.1 TLP e The attribute name for slot time needs to be spelled consistently with earlier uses in the standard.

Change to “aSlotTime” everywhere in this section


13.1.1.1 TLP e The A in CCA already stands for Assessment. You can’t have Clear Channel Assessment Assessment Time. Even

MS-Word flags it as redundant.

Change to “aCCATime” everywhere in this section


13.1.1.1 TLP e Missing paragraph mark after “aMACPrcDelay” Add end-of-paragraph mark after corrected “aMACProcessingDelay”


13.1.1.3 TLP t The concept of antenna appears, from all its occurrences, to be thought of within this standard as only relevant to RF. However, it could also apply to IR transmit/receive

apparatus

This applicability to IR should be pointed out somewhere within the text,

perhaps in one of the first antenna-related attribute definitions.

rejectedThere is no desire by the IR group to provide for multiple antennas

(Francois Lopez/Jan Boer)(9-0-2)

13.1.2 WD t Management objects are now defined twice: in the std body (section 13.1) and in Annex D. There is no added value in this double definition. Suggest to remove the definitions in the std body (13.1), if there is also a formal definitions in Annex D which has precedence anyway.

Suggest to use only one definition in the standard, which is to be

normative, and remove the other definitions.

One possibility is to remove the definition in the std body (13.1), and

postpone for full working group plenary

motion in full plenary to delete ANNEX D



#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


However the use of this MIB is primarily by the local MAC entity itself, and its use is not relevant for Network Management purposes This could be a good reason to specifically not place them in Annex D, but indeed specify them in section 13.1 The definitions per PHY as given in sections 14.8.2, 15.3.4 and 16.4 are considered very relevant, because they define the values for the attributes per PHY.

to correct Annex D as applicable.However a summary of the relevant

MIB parameters and their GET-REPLACE characteristics, like

provided in section 13.1.2 can be functional here, and could be

maintained in section 13.

A more clear alternative would be to maintain the section 13.1 definitions,

and remove them from Annex D, sinse these parameters are only of interrest to the lcal MAC entity.

Annex D removedby unanimous vote

13.1.4. RM e In the following subclauses, use consistent units1 should be in microseconds

13.1.4.12 RxRFDelay 13.1.4.13 aRxPLCPDelay13.1.4.15 aTxRampOffTime13.1.4.42 aHopTime13.1.4.44 aMaxDwellTime13.1.4.45 aCurrentDwellTime

comment accepted in part.Refer to comment resolution of

same comment in 14.8.2

13.1.4.1113.1.4.15

TLP t The time specified is an estimation of an actual future interval, and cannot be known exactly.

Change “The time in …” to “The nominal time in …”.


will also add reference to 9.2.3.1 for tolerance specifications

FG vote (11-0-1)13.1.4.

14RM T Y aMACPrcDelay is critical parameter, without a

defined value. Section 14.8.2.14 assumes a 2 usec value.

13.1.4.14 aMACPrcDelayMACPrcDelay ATTRIBUTE WITH APPROPRIATE SYNTAX integer; BEHAVIOR DEFINED AS "The nominal time in microseconds the MAC uses to process a frame and prepare a response to the frame"; aMACPrcDelay = 2usecsREGISTERED AS { iso(1) member-body(2) us(840) ieee802dot11(10036) phy(3) attribute(7) MACPrcDelay(14) };

comment withdrawn by submitter

13.1.4.15 TLP e Single occurrence of unknown unneeded acronym. Change “PA” to “Power Amplifier” accepted as result of accepting REVSEC10.DOC



#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


13.1.4.18 TLP e Inappropriate euphemism used, needlessly precludes use of this standard in space.

Change “over the air” to “through the wireless medium”


13.1.4.19 TLP T, E Yes This attribute is not a scalar, but a vector indexed by SID of all the other stations in the local BSS.

Please clarify your intent, or rewrite, or delete, or make this a structure with the MAC address or SID of the remote peer STA kept in the structure along with the

inter-station propagation time.

comment rejectedthere is currently no means to measure the propagation time.The full WG voted in a 1usec

fixed value some time ago(see also comment by MT in

general comment section)FG vote (11-0-2)

13.1.4.19 TLP t The time is anticipated, not known. This should be stated. Change to “The anticipated time it …” accepted as result of accepting REVSEC10.DOC

13.1.4.2 TLP The reader is unlikely to be familiar with the entire set of listed agencies. The countries corresponding to the agencies might be shown parenthetically. The list

terminator needs to be added to this set of values. Some formatting of the list, at least so that it commences on a

new line, would be useful.

Change to read “..the PLCP and PMD support in this implementation.

Currently defined values and their corresponding Regulatory Domains are:FCC (USA) = 10h, IC (Canada) = 20h, ETSI (most of Europe) = 30h, Spain =

31h, France = 32h, MKK (Japan) = 40h, list terminator = 00h";”.


13.1.4.20 TLP T The attribute is under-defined; it’s coding (other than intenger) is not specified and implementors from different countries would naturally make incompatible choices. For

example, is this coded as the minimum temperature of designed-for operation in milli-degrees Kelvin?

Add a specification of the attributes coding, either as a table of

corresponding ranges

acceptedtext will be added to elaborate on

the temperature ranges ofType 1 (0-40 deg C)

Type 2 (-20 - 55 deg C)Type 3 (-30 - 70 deg C)

corresponding changes will also be made to clauses 14, 15, and 16

PG vote (10-0-3)13.1.4.25 TLP e (1) It is unclear what is being measured or characterized

by this parameter. Is it a transmit FIFO and pipeline depth, or the number of bits per PHY symbol, or the

payload of an on-the-medium transmission unit, or what?

(2) The existing text is illiterate.

Clarify intent within the committee and rewrite appropriately, in literate English. For example, the existing text should be

rewritten to read “The maximum number of octets of an MPDU that can

be conveyed by a PLCPPDU”


13.1.4.2713.1.4.28

TLP e Failure in conceptualization. Surely antennae are not defined by integers. At least, not according to Webster’s

Rewrite each sub-sub-sub-section to a literate form, such as “Each antenna is



Tom Phinney, 01/03/-1,

The countries corresponding to the agencies might be shown parenthetically. Also, the list terminator needs to be added to this set of values.


#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


13.1.4.30 definition of “defined”. represented by an integer, starting with antenna 1, and through antenna N, where

N £255;”13.1.4.29 TLP e Poor exposition Rewrite as "This implementation's

support for diversity, encoded as: 01h — diversity is available and is

performed over the fixed list of antennas defined in aDiversitySelectionRx. 02h — diversity is not supported.

03h — diversity is supported and control of diversity is also available, in which

case the attribute aDiversitySelectionRx can be dynamically modified by the

LME."


13.1.4.4 WD E “Behaviour” not same as “Description” in Annex D. Suggest to remove the definitions in the std body (13.1), and to correct

Annex D as applicable.

comment acted on in reverse!as a result of WG motion to

delete ANNEX D13.1.4.4

4,13.1.4.45,7.3.2.

3,11.1.5,

14.8.2

SB t N Dwell time related MIB attributes are a complete mess in terms of units.

13.1.4.4 defines aMaxDwellTime and aCurrentDwellTime in nanoseconds (!), the default

values in 14.8.2 are in milliseconds and the comparison to a TSF timer value in 11.1.5 is to a time in

microseconds. Lastly the value for the dwell time in the FH Parameter set element (7.3.2.3) is in

Kmicroseconds.

Please can we have some order here. It would be nice if the aMaxDwellTime and aCurrentDwellTime were in Kus since this is what a number of other

MAC attributes such as aBeaconPeriod is in. It also ties up with the FH parameter set. It also

makes the TSF time comparison easy (hence the beacon stuff).

So:

aMAXDwellTime should be in Kus and be a default value of 390

(399.360ms)

aCurrentDwellTime should be in Kus an be a default value of 20.

defer to FH group for resolution

REFER to comment resolution of same comment in 14.8.2

13.1.4.55 TLP e Illeterate, perhaps partially due to typographic errors Rewrite as “This parameter, together with CCAWatchdogCountMax,

determines when energy detected in the channel can be ignored.”




#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


13.1.4.56 TLP e Illeterate, perhaps partially due to typographic errors Rewrite as “This parameter, together with CCAWatchdogTimerMax,

determines when energy detected in the channel can be ignored.”


13.1.4.all TLP e Many minor corrections are appropriate, as shown in the accompanying revision-marked files.

Change as shown in the accompanying revision-marked files.


13.all TLP E Yes Please take pity on non-native English speakers and use names that they have some slight chance of

understanding. Mis-pronounceable subsets of English words, such as “suprt” for “supported”, are not even close to acceptable. Similarly, what does “Asmnt” mean? How

about “Lvls”?

“Ths dcmnt is nt prntd fr clmns up.” That tried to say “This document is not printed four columns up.” Why are

vowels so scarce that you can’t use them? Please turn these names into something suitable for human

consumption. This clause is not acceptable as it stands. I am balloting NO on it, for gross inconsideration of the

intended readers.

To simplify the task of fixing this clause, I have applied global transforms to produce more intelligible attribute

names. See the submitted revision-marked files.

Use all of the letters in each constituent word unless the resulting word length is

really impractical.

See the submitted revision-marked files for an acceptable set of MIB names.


13.all14.all

TLP E Yes IEEE and ISO/IEC editing rules require use of SI units and proper nomenclature. That includes capitalizing a unit derived from a person’s name, and using the unit (W), not the name. It also includes using a non-break

space between the amount and the unit, so that line-wrap cannot split the amount from the unit

Follow the IEEE and ISO/IEC editng rules with regard to units; there is no

reason not to do so.

will use Kus instead of ms and will use SI units throughout the draft

14. JMZ E There are a number of uses of “is” that should be reworded as “shall” in the normative text of a standard.

Convert FH PHY English to IEEE Standardsese through clause 14.

comment accepted

14.2.2 RM T Y Clarify the supported data rates do not include all possible rates in the TXvector.

14.2.2 TXVECTOR ParametersThe following parameters are defined as part of the TXVECTOR parameter list in the PHY_TXSTART.request service primitive. The 1MBPS and 2MBS are the only rates currently supported. Other indicated data rates

adopt text changes(6-0-1)



#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


are for possible future use.14.2.2.1 TLP t A value of zero is non-sensical. How can the PHY be

asked to transmit nothing. The OSI Basic Reference Model does not permit null SDU transmissions, and there seems to be no reason for null PDU transmissions either.

Change minimum length from “0” to “1”.

comment acceptedchanged table to be consistent

(5-0-1)

14.2.2.2 vh e The FHSS MIB variable BSSBaicRate and the MIB variable CurrentHighSRate are mentioned here but are not defined in the respective clauses

Remove the last two sentences of the clause and insert: BASIC rate is 1.

HIGHSPEED is either 0 if not supported or 2 if the optional 2 Mbit/s PMD is implemented.

comment accepted.see resolution of comment by

RM below

14.2.2.2 RM t Y This section refers to undefined MIB variables 14.2.2.2 TXVECTOR PLCP_BITRATEThe PLCP_BITRATE parameter is an optional parameter. Its value describes the bit rate the PLCP should use to transmit the PLCP_PDU. Its value can be BASIC or HIGHSPEED. The BASIC rate is defined as the BSSBasicRate in the FHSS PHY MIB. The HIGHSPEED rate is defined by the CurrentHighSRate in the MIB.

Change text to read: The PLCP_BITRATE parameter

describes the bit rate at which the PLCP should transmit the

PLCPPDU. Its value can be any of the rates as defined in Error:

Reference source not found, and supported by the conformant FH

PHY.Naftali/Ron

7-0-014.3.1.1 TLP e The heading is missing all of its text. Add text to the heading line, or remove

the heading.comment accepted as result of

accepting REVSEC11.DOC14.3.1.1 TLP e “Function” is probably the maximally wrong word here.

FSM (finite state machine), procedure, automaton, etc. come to mind. But since function has a connotation of no or minimal side effects, it is probably not the best word to use. I don’t know what would be; perhaps the committee

can make that determination.

Choose a better word to convey the intended concept.

OPEN: FH editor to talk to commentor

14.3.1.12nd ¶

TLP e Arrows have orientation, and thus convey information which should be specified here.

Change to read “Each permissible transition between the states of a

function is represented graphically by an arrow from the initial to the terminal

state. A transition …”

comment accepted as result of accepting REVSEC11.DOC

14.3.2.1.1 TLP e poor English Change to read “… to detect a potentially-receivable signal, select …”


14.3.2.1.2 TLP E Yes Either the transmitted objects are “packets”, in which case this should be a “Start Packet Delimiter”, or they are

Use consistent nomenclature. Avoid the use of the term “packet” if possible,




#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


“frames”, in which case the word “packet” should be replaced by “frame” everywhere within this clause. Use

“packet” only if it refers to a PHY concept which must be distinguished from an 802.11 Data Link MAC PDU

(which latter is correctly called a “frame”).

because its primary meaning of “packet” is that of an OSI network-layer PDU,

and IEEE 802 has agreed to respect the OSI Basic Reference Model, including

its nomenclature.14.3.2.2.1 TLP e (1) A value of zero is non-sensical. How can the PHY be

asked to transmit nothing. The OSI Basic Reference Model does not permit null SDU transmissions, and there seems to be no reason for null PDU transmissions either

(2).LSB means Least Significant Byte, lsb means least significant bit. Its been this way for at least two decades.

Change to read “The PLCPPDU Length Word (PLW) is passed from the MAC as

a parameter within the PHYTXSTART.request primitive. The

PLW specifies the number of octets contained in the MPDU packet. Its valid

values are 001h - FFFh, representing counts of one to 4095 octets. The PLW is transmitted lsb first and msb last. The PLW is used by the receiving station, in

combination with the 32/33 coding algorithm specified in this clause, to determine the last bit in the packet.”

accepted by previous motion 14.2.2.1

14.3.2.2.2 TLP e (1) The table format should be corrected to fit within the column and avoid breaking the parameter name across

two lines.

(2) With regard to the spelled-out units, with one entry per line, clarity in this area might be worth more than the

paper saved.

See the submitted revision-marked files for the necessary corrections


14.3.2.2.3 TLP T Yes This polynomial works only when the modulation avoids differential coding, which has the effect of creating

double-bit errors on decoding. Otherwise two errors 22 bits apart can go undetected, as can many other low-

weight short error bursts, since the power of the CCITT code is biased heavily toward detecting odd numbers of

bits in error.

If differential decoding is required, then change to a CRC polynomial which does

not contain (1 + X) as a factor, so that the CRC polynomial is not compromised by the differential decoding process. No

change needed otherwise.

indeed the coding is not differential, so we do not have to

change.

14.3.2.3 TLP t The bit order must be specified, as well as the byte order. Change to read “… stream LSB and lsb first and MSB and msb last.“

Accepted change to msb/lsb. Deleted MSB/LSB since MAC determines octet

ordering.Naftali/Ron Unanimous

14.3.3 TLP e Poor terminology. Change the last two sentences to read “Execution of the PLCP state machines




LSB means Least Significant Byte, lsb means least significant bit. Its been this way for at least two decades, and so it shall remain.


See earlier annotation on the uselessness of requesting the transmission of nothing.


#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


normally is initiated by the FH PLME state machine and begins at the CS/CCA state machine. The PLCP returns to the FH PLME state machine upon interrupt to service a PLME service request, such

as PLMESET, PLMERESET, etc.”14.3.3.1.1 TLP T Yes In the Data Whitener Decoding Algorithm, the comment

/********* Calculate bias in header for format error checking ********

implies that there should be error checking. Where is it?

Add the necessary error checking procedure and any supporting text.

Deleted the reference to format error checking.

Ron/Carl7-0-0

14.3.3.2.1

SB t N In Figure 67 two timers are defined; count_down timer and CS/CCA timer. In this text/state machine CCA/CS timer has no actions other than ‘maintain’ - but there is no definition of what ‘maintain’ actually means. The accompanying text makes explicit reference to the purpose and actions on cown_down timer but only

makes rather vague references to ‘all relevant CCA/CS timers’ - there is only one such timer hinted at in the

state machine.

I could clearly take some sensible guesses here - but that does not make a good standard !

Make it clear what CCA/CS timers are required for compliance with the

standard (the comment author appreciates that much of the CCA

stuff is outside the scope of the standard). Now bring the state machine and text into line and

describe what the requirements and actions on the CS/CCA timer are.

Change CS/CCA text to read: If a PHYCCARST.request is received, the PLCP shall reset the CS/CCA

state machine to the state appropriate for the end of a

complete received frame. Delete other references to ÒCS/CCA

timersÓ.Carl/Ron

6-0-0

14.3.3.2.1


However, if the CS/CCA procedure indicates the start of a new frame within the countdown timer period, it is possible to transition to the receive procedure prior to the end of the countdown timer period. When a non-zero countdown timer reaches zero, the PLCP shall

reset all relevant CS/CCA assessment timers to the state appropriate for the end of a complete received frame

and the CS/CCA indication shall reflect the state of the channel.

This says that if I transition to a new frame within the countdown timer period then I keep the countdown

timer running from the previous frame and CCA locked busy until the countdown timer reaches zero (or is

updated).

Make intent clear in standard. Add statement: If the PHY transitions to receive under these conditions, the countdown timer shall be reset to the longer of (1) the remaining time of the current frame and (2) the length of the

new frame.Carl/Ron

5-0-1



#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


Was it the intent to have the countdown timer run and either expire during the new receive - or have an error in the new receive restart the timer. Alternatively, was

the intent to actually reset the countdown timer on entry into the new receive.

14.3.3.2.15th ¶

TLP e The wording “to the end as positively indicated” is very confusing; I can’t even figure out how it might be parsed

to make sense.

Rephrase to make the meaning clear. comment accepted by acceptance of REVSEC11.DOC

14.3.3.2.15th ¶

TLP e, t The wording “it is possible” is permissive as stated. If you wish to require such a transition, use “shall”.

Consider whether to make a requirement.

Comment accepted. Intended to be permissive rather than required.

Carl/Nathan6-0-1

14.3.3.2.2

9.2.5.2,12.3.5.1

0.2


The appropriate CS/CCA indication shall be generated prior to the end of each 50 µs slot time with the

performance specified in subclause Error: Referencesource not found (PMD).

(The CS/CCA indication is by PHYCCA.indicate as in figure 68)

While clause 12.3.5.10.2 says about PHYCCA.indicate:

This primitive shall be generated every time the status of the channel changes from channel clear to channel

busy or from channel busy to channel clear.

Clearly there is some conflict here - one says that the primitive is issued on a time basis once per slot time

even if the channel state has not changed, the other on a physical event (a change of channel state) irrespective of time. If I look at the PHY chapters the FH chapter

(Figure 68) would seem to follow 14.3.3.2.2 and the DS (Figure 83) follows 12.3.5.10.2

Actually this is pretty important for compliance given the rules that define when the back-off timer may, or

Correct conflict one way or the other - do I get a regular PHY CCA

indication per slot time, or only when the channel state changes.

(It also occurs to me that the first two sentences of clause 14.3.3.2.2 are

duplicated in the immediately previous clause.)

Addressed same comment in 12.3.5.10.2 in full PHY. Resolved by changing

ÒgeneratedÓ to ÒavailableÓ in 14.3.3.2.2.



#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


may not be decremented in 9.2.5.2 14.3.3.2.2

1st ¶TLP T It is not clear what “within a slot time including the PIFS

and DIFS windows” means. Does this mean that the slot time includes the PIFS and DIFS windows, or does it

mean a slot time plus a PIFS or DIFS window? Note the substantial difference in meaning depending on the way it

is actually worded.

Please clarify. Deleted reference to PIFS and DIFS window

14.3.3.2.22nd ¶

TLP e BRAVO!!! The word “perceived” is a great word choice. It conveys the ambiguity nicely.

None NONE

14.5.4.2 TLP e Last line of table. The word management is abbreviated as “mgmt”, not “mgnt”. The latter is an abbreviation for

“magnet”.

Change “PMD_PWRMGNT” to “PMD_PWRMGNT”, with any other

case and underscore changes as appropriate to match section 13.


14.5.4.3 RM t Y This section is inconsistent with 14.5.5.1 and 14.5.5.2. These sections already make provisions for support both

data rates using a common convention. If desired the 14.5.5.1 and 14.5.5.2 could be modified to allow

passing the BASIC and HIGHSPEED primitive within TXD_UNIT and RXD_UNIT.

TXD_UNIT PMD_DATA.request 1 Mbit/s: 0, 1 2 Mbit/s: 0, 1, 2, 3 RXD_UNIT PMD_DATA.indicate 1 Mbit/s: 0, 1 2 Mbit/s: 0, 1, 2, 3

comment withdrawn by commenter

14.5.5.4 TLP e This would be better titled “PA_RAMP”, rather than “PARAMP”. The first three times I read the word it

parsed par-amp, rather than p-a-ramp. Non-native English speakers will have even more difficulty.

Change “PMD_PARAMP” to “PMD_PA_RAMP”, with any other case and underscore changes as appropriate to

match section 13.


14.5.5.9 TLP e The term “power-saving” is used elsewhere in the standard for the function that is here referred to as “low-

power”.

Use the same terminology throughout the document; either choice is OK.


14.6.10 TLP E MS Word superscript and subscript font attributes produce unacceptable results.

Do not use MS-Word subscripting or superscripting; MS-Word makes the

resulting text TOO SMALL. Instead, select the characters to become the

subscript or superscript and use Format/Font/Font/Size/8 and

Format/Font/Character Spacing/Position/Lowered and

Format/Font/Character Spacing/By/2 for a subscript, and Format/Font/Font/Size/8

and Format/Font/Character Spacing/Position/Raised and




#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


Format/Font/Character Spacing/By/3 for a superscript.

(This is corrected in the submitted revision-marked files.)

14.6.14.4last ¶

TLP T Unclear relaxation of requirements. I believe that I know what is meant, but the existing wording would not stand up under legal scrutiny as a meaningful requirement, and

thus can’t be used as the basis for a conformance/nonconformance decision.

Please clarify this paragraph. Change wording to: An exception occurs when the total energy within a given 1 MHz channel as defined

by 14.6.5 exceeds the levels specified above.

Carl/Ron4-0-2

14.6.15.5

RM T Y The definition of Imp specifies that the desired signal amplitude is larger than the undesired. This makes no

sense as the specification is in + dB

Intermodulation protection (IMp) is defined as the ratio of the minimum amplitude of one of two equal interfering signals to the desired signal amplitude, where the interfering signals are spaced 4 and 8 Mhz removed from the center frequency of the desired signal, both on the same side of the center frequency.. desired signal strength to the minimum amplitude of one of two equal interfering signals at 4 and 8 MHz removed from center frequency, both on the same side of center frequency, that The Imp protection ratio is established at the interfering signal level that causes the FER of the receiver to be increased to 3% for MPDUs of 400 octets generated with pseudo random data, when the desired signal is -77 dBm. Each interfering signal is modulated with the FH PMD modulation uncorrelated in time to each other or the desired signal. The PMD shall have the IMp for the interfering signal at 4 and 8 MHz be greater than or equal to 30 dB.

Comment accepted.Carl/Ron

6-0-0



#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


14.6.15.6

PMK e FER used the clauses but not defned Insert in Clause 4: FER=Frame Error Rate (Vic Hayes: Ratio)

Comment accepted. Defined it at first appearance.

14.6.4 JMZ t By removing channel 47 from the Spain hop-sequences, it would be possible to come up with a single unified Spain/France table. I think it would be better to reduce the (potentially large) number of different regulatory domains that must be supported than to use all the possible frequencies in France.

Combine Spain/France into a single regulatory-domain.

Comment withdrawn by commentor. There are additional

country specific regulatory requirements outside the scope of

this standard.

14.6.8 JMZ t The mathematics behind the pseudo-random sequences should be explained so that (if one exists) a reverse-mapping function can be implemented. Trying to calculate what position in a hop-sequence a device is currently at requires a rather lengthy TSFTimer calculation or a sequential-search through the appropriate table. This makes predicting what frequency a STA will be on in the future (for Reassociation, for example) unnecessarily complex.

Explain the formula used to determine the hopping tables, or switch a formula with better mathematical properties (a number of academic articles on optimal patterns that pass regulatory muster have been published).

Commenter withdraws comment. He agrees that a statement in the informative annex explaining that the core patterns are generated by

a random number generator and filtered by an algorithm as

described in 95/246r1.

14.6.8 TLP t Specifications for France and Spain are made elsewhere, and need to be included here.

Change to read “p = number of frequency channels in hopping pattern (79 for North America/most of Europe,

23 for Japan, 11 for France, 9 for Spain)”

Accepted with changes: France is 27 and Spain is 35 channels.

Naftali/StuartUnanimous

14.6.8 TLP E The line formatting in this region leads to a difficult-to-read document, and the electronic version is very sensitive to the software set (OS, MS Word revision, font revision, selected printer, etc.) used for viewing. This sensitivity to

the reader’s environment is unnecessary.

Use the changed paragraph formatting provided in the submitted revision-marked files— don’t just put in line

breaks and manually wrap the lines. In other words, use MSWord the way

professionals do, not just as a flat-text program editor.


14.6.all TLP T In many places, specifications are made for Europe, and differently for France and Spain. The last time I checked,

France and Spain were in Europe. So all such specifications do not apply to “Europe” as claimed, but

only to “most of Europe”.

Change “Europe” to “most of Europe” wherever different specifications apply

to France or Spain.


14.7.2 RM e Missing “4” 14.7.2 4Level GFSK Modulation comment accepted14.7.2 TLP e Table 45 has incorrect title Change “Division” to “Deviation” comment accepted as result of

accepting REVSEC11.DOC14.8.2 SB t N Dwell time related MIB attributes are a complete mess Please can we have some order here. It Use SI units in all PHY



Last time I checked, France and Spain were in Europe. So this large hopping set does not apply to “Europe” as claimed, but only to “most of Europe”.


#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


7.3.2.3,11.1.5,

13.1.4.44,

13.1.4.45,

in terms of units.

13.1.4.4 defines aMaxDwellTime and aCurrentDwellTime in nanoseconds (!), the default

values in 14.8.2 are in milliseconds and the comparison to a TSF timer value in 11.1.5 is to a time in

microseconds. Lastly the value for the dwell time in the FH Parameter set element (7.3.2.3) is in

Kmicroseconds.

would be nice if the aMaxDwellTime and aCurrentDwellTime were in Kus since this is what a number of other

MAC attributes such as aBeaconPeriod is in. It also ties up with the FH parameter set. It also

makes the TSF time comparison easy (hence the beacon stuff).

So:

aMAXDwellTime should be in Kus and be a default value of 390

(399.360ms)

aCurrentDwellTime should be in Kus an be a default value of 20.

parameters. The minimum unit of time will be ms.

Stuart/CarlFH PHY: 5-0-1

Plenery motion:Use Kms rather than ms.

WG: Bob/Johnny Passes with 1 opposing

Change 400 ms to 390 Kms and 20 ms to 19 Kms in 14.8.2 and 14.8.2.1.37 and 14.8.2.1.38.

FH: Ron/George5-0-1

14.8.2 RM t N The default values for Cwmin and Cwmax are incorrect.

aCWmin 15 decimalh aCWmax 1023 decimalh

Comment accepted.Ron/Carl

3-0-014.8.2 TLP E Use of term “Dep” in final column. If you wish to use a

shortened form that fits on a single line, then choose one that is meaningful to non-native-English speaking readers and explain it in the Notes which follow the table, as in

“where Implementation means that the behavior is dependent on the specific implementation”.

Use an appropriate legitimate word, or add an explanatory note to the table.


14.8.2.1.18

TLP e The Symbol font contains a multiply character “´”; use it, rather than the letter “x”.

Use the correct character for multiplication.


14.8.2.1.2 TLP E Yes The reader is unlikely to be familiar with the entire set of listed agencies. The countries corresponding to the

agencies should be shown.

Add a third column to the table specifying the region/countries to which

each code point applies.

(This is shown in the submitted revision-marked files.)


14.8.2.1.4 TLP t or e The value assigned to the attribute is not equal to the value computed from the formula which defines the

attribute. The formula gives 27 + 20 +1 = 48, not the claimed 50. If you intend that the number should be rounded up by including a safety factor, then say so.

Correct something. Change equation in 14.8.2.1.4 to be: aCCATime +

aRxTxTurnaroundTime + aAirPropagationTime +

aMACProcessingDelay to be



#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


Wording such as that found in the definition of aSIFSTime would be acceptable. But claiming equality

without making the sums match is not acceptable.

consistent with 9.2.9. Change to MAC figure in 9.2.9 to

subtract aRxRFDelay and aRxPLCPDelay from aCCADelay to = aCCATime. Leave aSlotTime

at 50 us. Add clarification in 13.1.4.5 that aCCATime includes aRxRFDelay and aRxPLCPDelay.

FH: Carl/Ron 7-0-0PHY: Carl/Al 6-0-3

14.all TLP E The earlier clauses in the document do not use an underscore after the prefix PHY or PLCP, or PLME, even though that might aid readability. So this clause should

not either. The necessary corrections have been included in the submitted revision-marked files, but the figures

have not been corrected.

Be consistent throughout the draft standard – either use hyphens or

underscores, which would improve readability, or don’t. But do so

consistently.

change accepted as result of accepting REVSEC11.DOC



about “Lvls”?




intended readers.



Make names consistent with the name changes made in section 13 as a result of

the similar comment for section 13

Also update the figures, which I was not able to do in the submitted revision-

marked files.

change accepted as result of accepting REVSEC11.DOC

14.all5.1.1.2 (c)

5.2.4.15.4

9.2.1


transmission.


accepted with acceptance of REVSEC10.DOC

andREVSEC12.DOC



#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


12.all15.some

16.all15 MT E in order to maintain consistency with other sections,

the DSSS section should have added the France and Spain regulatory domains.

Updates to 15.4.6.2, 15.3.2, 15.3.3.3, DSSS PICS, MIB description.

Other editorial fix-ups provided in separate file

This text was provided in a previous comment. France allows operation

from 2.4465 to 2.4835 GHz (4 channels possible). Spain allows

operation from 2.445 to 2.475 GHz (2 channels).

Two additional channels could be added to the DSSS channel plan for ETSI and France (2467 and 2472)

acceptedwith SEC12.DOC as editorial

changes

acceptedadd upper two channels as well as lower two channels to ETSI

and upper two channels to France

15.1.1last ¶

TLP e This paragraph is inappropriate as worded. It sound more like instructions to a standards-writing committee than the

finished output of that committee. Either remove it or restate it as accomplished fact, rather than hypothetical necessity. Also, there is only one PMD sublayer in your mode, so there can be only one in this clause (perhaps

with variations). So what does the first sentence mean? This is just sloppy writing, in my opinion.

Clean up this paragraph or remove it. Accepted as result of accepting REVSEC12.DOC

15.1.3 PMK e Additional subsets of acronyms are introduced Consolidate 15.1.3 with 4 to have just one table of acronyms

comment accepted as result of accepting SEC12.DOC

15.1.3 4

MT e add the abbreviations from clause 15 (DSSS PHY) this maintains consistency among clauses

add abbreviations from clause 15 and delete from clause 15


15.2.3.6 DSM t I do not see how a 32 byte MPDU can be transmitted in 192 microseconds(assuming a transmission rate of

1 Mbps)

Change to 256 microseconds point takenresolution will be to change the 32 byte reference to 24 bytes - this is treated as an editorial

change15.2.6 PMK e “PLCP transmit procedure is shown in figure 6”. “procedures is shown in figure 81”. comment accepted as result of

accepting SEC12.DOC15.3,4 PMK e “...specific values defined in Table 3.” “defined in Table 58”. comment accepted as result of

accepting SEC12.DOC 15.3.1 PMK e “Table 1 lists this primetives....” “Table 56 lists the primitives” comment accepted as result of

accepting SEC12.DOC



#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


15.3.4 SB e N It says here ‘All DSSS PHY Layer MIB attributes are defined in clause 12 with specific values defined in

table 3’ Table 3 is Duration/ID Field Encoding - this should be a

reference to the following table (Table 58 in D5).

The title on Figure 58 is ‘MIB Variable Parameters’ whereas I believe it should more accurately be titled

‘MIB Attribute Default Values/Ranges’

Correct reference and title as suggested.

The text that appears underneath Table 47 (FHSS PHY Attributes)

relating to the meaning of static/dynamic could also be reproduced here for clarity.


15.3.4 p.243

WD e Reference to “clause 12” should be “clause 13.

The contents of this table does not match the contents and sequence of the applicable groups as defined in Annex D, and or section 13.1.2

Suggest to remove the definitions in the std body (13.1), and to correct

Annex D as applicable.

editorial reference correction accepted

ANNEX D deleted by WG motion

15.4.6.2 AK T Yes Reduce the number of defined channels for FCC and ETSI domains.

Channels can not be used in the same area because they (heavily) overlap. Adjacent channel rejection is 35 dB with 30 MHz spacing (15.4.8.3).Definition of this many channels does not improve network performance but makes channel allocation and channel acquisition (handover/roaming/start up) more complex. Define only 3 channels: preferably 2422, 2444 and 2466 for both FCC and ETSI and adapt table 63 (and appendix A.4.6) accordingly. Adapt table 63 accordingly. (also appendix A.4.6 is to be adapted)

comment rejected

The proposed channel plan reduces the interference

avoidance capabilities of the DS PHY. Further, there are ample

ways to determine the operational frequency by

utilizing a management function which operates above the MAC

layer. The draft does not support DS channel mobility

and therefore the ‘auto’channel recognition capability described is currently beyond the scope of

the standard.In terms of making the channel acquisition more difficult, with

the addition of DSSS PHY elements in the beacons (which

are proposed in other comments) the aquisition

uncertainty becomes a non-issue



#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


entirely.

(6-0-0)15.4.6.2 AK T Yes Make channel 1 and 2 optional for FCC and IC With the current channel definition it

is not possible to manufacture a product that is FCC/IC compliant and ETSI compliant and IEEE compliant (three labels on the same device). With channel 1 and 2 optional such a device is possible (if it actually does not support channel 1 and 2).Advantage: same product for both American and European market.Disadvantage: in a network in FCC domain operating on channel 1 or 2 an ETSI/FCC device can not have a connection.

comment rejected

allowing for optional channels gives way to interoperability

issues

The addition of two lower channels to the ETSI domain were added to match the FCC

domain. However, the addition of two upper channels were also

added to the ETSI domain which does not allow for the

resolution of this comment for this causes the same situation.

It is felt that the additional spectrum is more worthwile

than the striving for a common product.

motion to accept this resolutionapproved by (5-0-1)

15.4.6.4 PMK e “DBPSK encoder is specified in Table 9”. “is specified in Table 64” comment accepted as result of accepting SEC12.DOC

15.4.6.4 PMK e “DQPSK encoder is specified in Table 10”. “is specified in Table 65”. comment accepted as result of accepting SEC12.DOC

15.4.7.1 PMK e “regulatory bodies is shown in Table 11”. “is shown in Table 66”. comment accepted as result of accepting SEC12.DOC



about “Lvls”?

“Ths dcmnt is nt prntd fr clmns up.” That tried to say


the similar comment for section 13. Also update the figures.




#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


“This document is not printed four columns up.” Why are vowels so scarce that you can’t use them? Please turn

these names into something suitable for human consumption. This clause is not acceptable as it stands. I

am balloting NO on it, for gross inconsideration of the intended readers.

15.some 5.1.1.2 (c)

5.2.4.15.4

9.2.112.all14.all

16.all


transmission.



REVSEC12.DOCand manual change to clause 16

16.1.1last ¶

TLP e This paragraph is inappropriate as worded. It sound more like instructions to a standards-writing committee than the

finished output of that committee. Either remove it or restate it as accomplished fact, rather than hypothetical necessity. Also, there is only one PMD sublayer in your mode, so there can be only one in this clause (perhaps

with variations). So what does the first sentence mean? This is just sloppy writing, in my opinion.

Clean up this paragraph or remove it. comment acceptedparagraph removed

16.2.1 PMK e prepended if an English word is a rare and obscure one “a PLCP Preamble and PLCP Header are added to the MPDU...

comment accepted

16.2.4.12nd ¶

TLP e, t The phrase “transitions in L-PPM slots which would otherwise constitute an illegal symbol” which ends this

paragraph has not been defined.

Either describe here what you mean, or add a forward reference to the (sub)^N-

clause where these concepts are described.

Comment rejected.Legal symbols are defined in

tables 67 and 68

16.2.4.216.2.4.316.2.4.4

PMK e “The SFD field is not modulated using4-PPM but instead consists of transitional in 4-PPM slots which would otherwise constitute an illegal symbol”. This is completely incomprehensible. what is the otherwise illegal symbol?

Comment rejected.Legal symbols are defined in

tables 67 and 68

16.2.4.5 TLP e (1) The normal computer convention is “lsb” and “msb” refer to bits, “LSB” and “MSB” refer to Bytes.

(2) If the qualifier “in time” is needed here, then it is

Change to read “The lsb (least significant bit) shall be transmitted

first.”

comment accepted



#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


needed at all earlier occurrences of “shall be transmitted first”. “in time” seems redundant. How can it be

transmitted first, yet not be first in time?16.2.4.6 PMK e Is the ones compliment of the remainder. Is the ones complement of the

remainder.comment accepted

16.2.5.all

TLP e Use “shall” rather than “will” when the intent is legislative. See the submitted

revision-marked files.

comment accepted

16.2.5.2 TLP e, t Specify the range of application of the data rate. See the submitted revision-

marked files.

comment accepted

16.2.5.3 TLP e, t Specify that reception is the relevant process. See the submitted revision-

marked files.

comment accepted

16.3 TLP t to ??? To what is the PHYSAP presented? Add the missing destination. comment rejectedPHYSAP is defined in 12.2

16.3.2.1 TLP e, t Inadequate lead-in to table 67 Change last sentence of first paragraph to read “Transmission order of the symbol slots is from left to right, as

shown below, where a 1 indicates in-band energy in the slot, and a 0 indicates

the absence of in-band energy in the slot”

comment accepted

16.3.3.1table 69

PMK e Collumn Heading= Peak Optical Power (averaged over the pulse width) Is it peak power or average power?

Peak Optical Power (over the pulse width) if this is what is meant

comment accepted“peak value” is used

16.3.3.2 TLP e If you prefer the “xx than or equal to” form of expression, then use “less”, not “lower”, since numeric comparison, and not height in a gravitation field, is being discussed.

Correct the text to reflect intended meaning.

comment accepted

16.3.3.2and

following

TLP E Yes IEEE and ISO/IEC editing rules require use of SI units and proper nomenclature. That includes capitalizing a unit derived from a person’s name, and using the unit (W), not the name. It also includes using a non-break

space between the amount and the unit, so that line-wrap cannot split the amount from the unit

Follow the IEEE and ISO/IEC editng rules with regard to units, including time units (s, ms, ms, ns, ps, fs, etc.) ; there is

no reason not to do so.

comment accepted

16.3.3.3 TLP e Correct the formatting of Table 71 as shown in the submitted revision-marked files.

Make the table less than the full column, with the heading Bold as in the previous

table, as shown.

comment accepted

16.3.3.3 TLP e The statement “may be added at a future time” is not Replace with “are for future study” comment accepted



#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


acceptable in a standard.16.3.5.1 TLP E Neither CS or ED have been described to this point, nor

have any mechanisms or models of operation been proposed by which a reader could infer what CS and ED

imply. And the generic names CarrierDetect and EnergyDetect do not convey enough information about

the nature of the detection process or the implied hardware to permit the reader to continue attempting to

understand this clause.

Please propose a model of receiver operation before referring to the

behavior of the model’s constituent parts.

<addendum after reading the CS and ED descriptions> I strongly recommend that

this ordering problem be remedied by describing ED first, then CS, and then CCA. Had this been done in the draft,

this comment would never have existed.

comment accepted

16.3.5.2 TLP e The second sentence is redundant; it is better placed where it occurs later in the sub-sub-sub-clause, at the end.

Remove the second sentence; it is 100% redundant.

comment accepted

16.4 TLP e Table 73, rows for aMPDUMaxLengthXX.

Section 10 lists a single attribute aMPDUMaxLength, not a number of data-rate-dependent attributes. One of these

lines needs to be struck, as shown in the submitted revision-marked files.

Delete the data rate from one label, and the second row with the same label

prefix.

comment accepted



about “Lvls”?




intended readers.




the similar comment for section 13.

Also update the figures, which I was not able to do in the submitted revision-

marked files.

comment accepted

16.all TLP e The wrong prefix is used with PDU and SDU. Replace PPDU and PSDU with comment accepted



#Section number

your voter’

s id code

Cmnt typeE, e, T, t

Part of

NO vote


PLCPDU and PLCSDU as appropriate. (Replacements made in submitted

revision-marked files.)16.all

5.1.1.2 (c) 5.2.4.1

5.49.2.112.all14.all

15.some


transmission.


for clause 14 ,15,16 accepted as result of accepting

REVSEC11.DOC and REVSEC12.DOC

and correction made to clause 16

Figure 84

DSM t There are state transition lines in the figure that go nowhere.

Add connections to the lines so that the two floating lines at the lower

right of the figure connect with the line in the upper right.


Documents

IEEE 802.11/D2 Letter Ballot Response€¦ · Web viewResolutions for Comments on Clauses 12-16. 12.3.2 PMK e “...is seperated into to sublayers:...” “is separated into two